LOGIC AND CONVERSATION*
It is a commonplace of philosophical logic that there are, or appear to be, divergences in meaning between, on the one hand, at least some of what I shall call the FORMAI,devices--, A, V, 3, (x),3(x),S x (when these are given a standard two-valued interpretation)-and, on the other, what are taken to be their analogs or counterparts in natural language - such expressions as not, and, or, if, ~11, some (or at least one), the. Some logicians may at some time have wanted to claim that there are in fact no such divergences; but such claims, if made at all, have been somewhat rashly made, and those suspected of making them have been subjected to some pretty rough handling. Those who concede that such divergences exist adhere, in the main, to one or the other of two rival groups, which for the purposes of this article I shall call the formalist and the informalist groups. An outline of a not uncharacteristic formalist position may be given as follows: Insofar as logicians are concerned with the formulation of very general patterns of valid inference, the formal devices possess a decisive advantage over their natural counterparts. For it will be possible to construct in terms of the formal devices a system of very general formiilas, a considerable number of which can be regarded as, or are closely related to, patterns of inferences the expression of which involves some or all of the devices: Such a system may consist of a * U s e d hy p e r m i s s i o n o f t h e allthtir and pllhlisher f r o m Il. James Lectures, delivered at I
llarvard
I;trwrd Ilniversity Prew. (;opyrigl\t
Ilniversity
Pal11
in 1967, and to
Griw’s
Willi;lrn
1x1 pllhlishcd
hy
I975 I)y I I . P~III Grice.
41
42
11. P.
Crice
lliI\~c the nlmniiig that II:IS lxen assigilcd to them, a11d ai1 indefinite nIlrnl>er of’ frlrthcr fi,rinltl;~s, ni;l~iy of them less ohviollsly acceptable,
tach of wllicll can he shown to l,e acceptable if the memljers of the original set are ac*ceptal)le. WC leave, tlltts, a w a y o f hat~dling cliil)iollsly ;lcceptal,le patterns of inference, and if, as is sometimes possilJe, we c a n a p p l y a decision profdllre, we have an even better way. Fllrthcrmore, from a philosophical point of view, the possession I)y the natliral collnterparts of those elements in their meaning, which they do not share with the corresponding formal devices, is to he regarded as an imperfection of natural languages; the elements in cltlestion are \lndesiral,le excrescences. For the presence of these t>lcments leas the restlIt tllat the concepts within which tllcy appear callnot I)e precisely/clearly defined, and that at least some statements invol\ring them cannot, in some circumstances, be assigned a definite trlltli vnlllc: x11(1 tlic indefiniteliess of these concepts is not only ohjc~tional~lc in itself l,lit leaves open the way to metaphysics- we cannot I,e certain tliat none of these natural language expressions is In~~t~~pllysic.nlly ‘loaded’. For these reasons, the expressions, as used in natural speech, cannot I)e regarded as finally acceptaljle, and may turn ollt to I)e, finally, not fully intelligible. The proper course is to conceive and hegin to constrrlct an ideal language, incorporating the formal devices, the sentences of which will be clear, determinate in truth value, and certifiably free from metaphysical implications; the foundations of science will now he philosophically secure, since the statements of the scientist will be expressible (though not necessarily actually expressed) within this ideal language. (I do not wish to suggest that all formalists would accept the whole of this outline, but I think that all would accept at least some part of it.) To this, an informalist might reply in the following vein. The philosophical demand for an ideal language rests on certain assrlmptions that shorlld Ilot he concedd; these are, that the primary yardstick by which to jrldge the adequacy of a language is its ability to serve the needs of science, that an expression cannot he guaranteed as fully intelligible tInless a n explication or analysis of its meaning has been provided, and that every explication or analysis must take the form of a precise definition that is the expression/assertion of a logical equivalence. Langtlage serves many important purposes besides those of scieiitific incluiry; we can know perfectly well what an expression means (and so a fortiori that it is intelligible) without knowing its analysis, ant1 the provision of an analysis may (and usually does) consist in the specification, as generalized as possible, of the conditions
Logic
and
Conversation
4.3
that colrnt for or against the appIicaI)iIity of the expressio~t Iwing analyzed. Moreover, while it is no dorlbt true that the fi)rllliil (1cviccs are especially amenable to systematic treatment by the Iogiciall, it remains the case that there are very many inferences and argirments, expressed in natural language and not in terms of these devices, that are nevertheless recognizably valid. So there must Ix a placct fi)r a11 unsimplified, and so more or less unsystematic, logic of the natural counterparts of these devices; this logic may Ire aided and guided Iry the simplified logic of the formal devices hut cannot I,e srlpplanted by it; indeed, not only do the two Iogics differ, Irrit sometimes they come into conflict; rules that hold for a formal device may not hold for its natural counterpart. Now, on the general question of the place in philosophy of the reformation of natural language, I shall, in this article, have nothing to say. I shall confine myself to the dispute in its relation to the alleged divergences mentioned at the outset. I have, moreover, no intentiorr of entering the fray on behalf of either contestant. I wish, rather, to maintain that the common assumption of the contestants that the divergences do in fact exist is (broadly speaking) a common mistake, and that the mistake arises from an inadequate attention to the nature and importance of the conditions governing conversation. I shall, therefore, proceed at once to inquire into the general conditions that, in one way or another, apply to conversation as such, irrespective of its subject matter.
IMPLICATURE
Suppose that A and B are talking about a mutual friend, C, who is now working in a bank. A asks B how C is getting on in his job, and B replies, Oh quite well, 1 think; he likes his colleagues, (~no! he husn’t been to prison yet. At this point, A might well inqriire what I3 was implying, what he was suggesting, or even what he meant by saying that C had not yet been to prison. The answer might be any one of such things as that C is the sort of person likely to yield to the temptation provided by his occupation, that C’s colleagues are really very unpleasant and treacherous people, and so forth. It might, of course, be quite unnecessary for A to make such an inquiry of B, the answer to it being, in the context, clear in advance. I think it is clear that whatever B implied, suggested, meant, etc., in this example, is distinct from what B said, which was simply that C had not been to prison yet. I wish to introduce, as terms of art, the verb implircrtc and
44
11. P.
Crice
tile r e l a t e d n011ns inlj~lictrt~rw ( c f . iufl~lyir~g) a n d ir,lj)lic-cltfrnl ( c f . is ilrr~~licd). The point of this manetlver is to avoid leaving, on each occasion, to choose between this or that meml)er of tile family of verbs for which itn),liccltr is to do general ditty. I shall, for the time being at least, have to ;\sslllne to a considernl)lc clxtcnt an intllitive ul~clerstanding of the incalling of ,~(I!] in suc‘h contexts, and an ability to recognize particlllnr vcrhs as members of’ tllc family with which itnj,lir*c/te is associated. I can, however, make one or two remarks that may help to clarify the more prol,lematic of these assumptions, namely, that connected with the meaning of the word OS0 y. In the sense in wliicll J am using the word say, 1 intend what someone has said to be closely related to the conventional meaning of the words (the sentence) he 11~s uttered, Suppose someone to have uttered the sentence Ilc is in tllp grill c!fn uic.4. Given a knowledge of the English language, l)tlt no knowledge of the circllmstances of the rltterance, one would know something al)ollt what the speaker had said, on the assllmption that he was speaking standard English, and speaking literally. One would know that he had said, al>out some particular male person or animal .x, that at the time of the rltterance (whatever that was), either (1) x was tlnable to rid himself of a certain kind of l>ad character trait or (2) some part of x’s person was caught in a certain kind of tool or instrtlment (approximate account, of course). But for a fIlli identification of what the speaker l\i\cl said, 011c would need to know (a) the identity of X, (1,) the time of utterance, and (c) the meaning, on the particular occasion of utterance, of the phrase in tile grill c!f’cl oicle [a decision ljetween (1) and (Z)]. This brief indication of my use of s(i!j leaves it open whether a man who says (today) I1rrrold Wilson is (I great man and another who says (also today) TIze British Prime Minister is (1 greclt nzcln would, if each knew that the two singular terms had the same reference, have said the same thing. But whatever decision is made about this question, the apparatus that I am al)out to provide will be capallle of accollnting for any implicatures that might depend on the presence of one rather than another of these singular terms in the sentence uttered. Sllch implicatures wor~lcl merely be related to different maxims. In some cases the conventional meaning of the words {Ised will determine what is implicated, hesides helping to determine what is said. If I say (smugly), He is an Englishmntl; he is, thewfhre, hoe, I have certainly committed myself, 1)~ virtue of the meaning of my words, to its Ijeing the case that his being ljrave is a conseqllence of (follows from) his being an Englishman. But while I have said that tL\h(/f
L o g i c and C o n v e r s a t i o n
45
he is an Englishman, and said that he is brave, I do not want to say that I have SAlI) (in the favored sense) that it follows from his being an Englishman that he is brave, thoirgh I have certainly indicated, and SO implicated, that this is so. I do not want to say that my irtterante of this sentence worild l)e, s’IxI(:*I’I,\’ SPKAKINC:, false sho~~ld the c!onscY~rlence in qriestion fail to hold. so SOME implicatures are conventional, unlike the one with which I introduced this discussion of implicature. I wish to represent a certain s111~class of nonconventional implicatures, which I shall call (:ONVI~I~SA~‘IONAI, implicatlires, as being essentially connected with certain general features of disc*oiirse; so my next step is to try to say what these features are. The following may provide a first approximation to a general principle. Our talk exchanges do not normally consist of a succession of disconnected remarks, and would not he rational if they did. They are characteristically, to some degree at least, cooperative efforts; and each participant recognizes in them, to some extent, a common purpose or set of purposes, or at least a mutually accepted direction. This purpose or direction may he fixed from the start (e.g., by an initial proposal of a question for discussion), or it may evolve during the exchange; it may he fairly definite, or it may he so indefinite as to leave very considerable latitude to the participants (as in a casual conversation). But at each stage, SOME possible conversational moves would be excluded as conversationally unsuitable. We might then formulate a rough general principle which participants will be expected (ceteris paribus) to observe, namely: Make your conversational contribution such as is required, at the stage at which it occurs, 1,~ the accepted purpose or direction of the talk exchange in which you are engaged. One might label this the COOPEHATIVE PRINCIPL,E.
On the assumption that some such general principle as this is acceptable, one may perhaps distinguish four categories under one or another of which will fall certain more specific maxims and submaxims, the following of which will, in general, yield results in accordance with the Cooperative Principle. Echoing Kant, I call these categories Quantity, Quality, Relation, and Manner. The category of (&JAN?‘I?‘Y relates to the quantity of information to be provided, and under it fall the following maxims: 1. Make your contribution as informative as is required (for the current purposes of the exc:hange). 2. Do not make your contribution more informative than is required.
FF. P.
C.&e
(The (;ccontl maxim is disprital)le; it miglit be said that to he overinformative is not a traiisgrctssion of the CP hut merely a waste of time. IIowever, it miglit IX answ~~rc~~l t h a t siich overinform;ttivclless may lie conhising in that it is lialjle to raise side issrres; and there may also l)e a11 indirect effect, in that the hearers may he misled as a rc>s\ilt of thinking tlrat thcrtl is some particiilar 1~OINl’ in the provision of the excess of information. Ilowever this may be, there is perhaps a different reason for rloriht aborit the admission of tliis second maxim, namely. that its effect will he secrired lry a later maxim, which concerns relevance.)
I. Do not say what you believe to he false. 2. DO not say that for which you lack adequate
evidence.
IJnder the category of ~~EIATION I place a single maxim, namely, ‘Be relevant.’ Thorigh the maxim itself is terse, its formulation conceals a nrimber of problems that exercise me a good deal: questions al~ilt \vhat different kinds ant1 focirses of relevance there may be, how these shift in the corirse of a talk exchange, how to allow for the fact that SI I1 ejects of conversation are legitimately changed, and so on. I find the treatment of sr~ch cliicastions exceedingly difficrilt, a n d I Irope to revvert to tllem in a later work. Finally, under the category of AIANNEH, which I trnderstand as relating not (like the previous categories) to what is said but, rather, t o HO\\’ what is said is to be said, I include the supermaxim-‘Be perspicuous’ - and various maxims such as: 1. Avoid ol)scririty of espression. 2. Avoid amhigu i ty. 3. Be brief (avoid unnecessary prolixity). 4. Be orderly. And one might need others. It is obvious that the observance of some of these maxims is a matter of less urgency than is the observance of others; a man who has expressed himself with undue prolixity would, in general, be open to milder comment than would a man who has said something he believes to he false. Indeed, it might be felt that the importance of at least the first maxim of Quality is such that it shorrld not be incliided in a scheme of the kind I am constructing; other maxims come into operation only on the assumption that this maxim of Quality is satisfied. Wliile this may he correct, so far as the generation of
I,ngic
and Conversation
47
implicatures is concerned it seems to play a role not totally different from the other maxims, and it will Ire convenient, for the present at least, to treat it as a member of the list of maxims. There are, of course, all sorts of other maxims (aesthetic, social, or moral in character), such as ‘Be polite’, that are also normally observed by participants in talk exchanges, illld these may also gcncratc nonconventional implicatures. The conversational maxims, lrowcvcr, and the conversational implicatures connected with them, arc spccially connected (I hope) with the particrrlar prrrposcs that tirlk (and so, talk exchange) is adapted to serve and is primarily enrployed to serve. I have stated my maxims as if tlris ptrrposc were a nrnximally effective exchange of information; this specification is, of’ colrrsc, too narrow, and the scheme needs to be generalized to allow for srrch general purposes as influencing or directing the actions of others. As one of my avowed aims is to see talking as a special case or variety of purposive, indeed rational, behavior, it may be worth noting that the specific expectations or presumptions connected with at least some of the foregoing maxims have their analogues in the sphere of transactions that are not talk exchanges. I list briefly one such analog for each conversational category. 1. Quantity. If you are assisting me to mend a car, I expect yor11 contribution to be neither more nor less than is reclrrircd; if, for example, at a particular stage I need four screws, I expect you to hand me four, rather than two or six. 2 . Q u a l i t y . I expect your contributions to be genuine and not spurious. If I need sugar as an ingredient in the cake you are assisting me to make, I do not expect you to hand me salt; if I need a spoon, I do not expect a trick spoon made of rubber. 3. Relation. I expect a partner’s contribution to be appropriate to immediate needs at each stage of the transaction; if I am mixing ingredients for a cake, I do not expect to be handed a good l)ook, or even an oven cloth (thorrgh this might be an appropriate contribution at a later stage). 4. Manner. I expect a partner to make it clear what contribution he is making, and to execute his performance with reasonable dispatch. These analogies are relevant to what I regard as a fundamental question about the CP and its attendant maxims, namely, what the basis is for the assumption which we seem to make, and on which (I hope) it will appear that a great range of implicatures depend, that talkers will in general (ceteris paribrrs and in the absence of indica-
tionc t o the contrary) proceed irj the ri~anner~ t h a t tllclsc princiiples presc~riI,c~. A (11111 I,rlt, tro dorrl)t :\t ;I certain level, uleclrtatc~ ;~r~swcr is that it is jtrst a well-reu~gnizctl c~rnpiric21 Fact that ptq3lc~ I)0 hehave in these ways; they h:~ve lenriicd to do so in c~hildl~ootl :irld not lost tl)C? habit Of doing SO; ilJl(1, in(lc‘e(l, it \Y011l(l iljvolvc> it g0Otl tlCb;il Of Cffort to make a radicd dcpartttre from the hal)it. It is rnt~ch easier, fol exitmple, to tell the trrlth than to invent lies. I am, however, eno11g11 of a r:~tionalist to wtlnt to find ;I ljasis that rlnderlies these facts, r~ndc~~rial~le tho~rgh they may 1~; J worrl~l like to 1~ able to think of the stnn(\;~rd type of c(~rlvers~~tiorl;ll prac%ce not rrierely as sornetliirig that all or most do IN FA(:‘l’ follow Ijut as sometlliilg t h a t it is Rl+~SoNAI3I,r~: for 11s t o follow, t h a t WV Sllol~I,l> NOT al)ancloll. For a time, I was attnMed by the iden tliitt 0l)servnnce of the CP and the maxims, in a talk exchange, co~rld be thought of as a clri~~si-contractrlal matter, with parallels outside the realm of discorrrse. If YOII pass hy when I am strrlggling with my stranded car, I no dorrbt have some degree of expectation tlrat you will offer help, hlrt once yo11 join me in tinkering \lnder the hood, my expectations become stronger and take more specific forms (in the al,sence of indications that you are merely an incompetent rnecldler); and talk exchanges secbnicd to me to exhibit, cliar;lcteristic~ally, cert:lili features that jointly distinguish cooperative transactions: 1. The put-ticipallts have some common immediate :\im, like getting a car mended; their ultimate aims rnay, of course, he independent ancl even in conflict-each may want to get the car mended in order to drive off, leaving the other stranded. In characteristic talk exchanges, there is a common aim even if, as in an over-the-wall chat, it is a second-order one, namely, that each party should, for the time l>eirig, identif;/ himself with the trailsitory coi~vei.s~~tion~~l interests of the other. 2. The contributions of the participants sho111d IX: dovetailed, mrrtually dependent. 3. There is some sort of understanding (which may 1)e explicit but which is often tacit) that, other things being eqld, the transaction shorlld continue in appropriate style [Inless both parties are agreeable that it shor~ld terminate. You do not jrrst shove off or start doing something else. HIIt while some s~icli quasi-contractual l)asis as this rilay a p p l y t o some cases, there are too many types of exchange, like quarreling and letter writing, that it fails to fit comfortably. In any case, one feels that the talker who is irrelevant or ol)sctlre has primarily let
49 down not his audienct~ l,llt himwlf. S o I wo~~ld like to lx: ill)le to show that observance of lll (axpected to have an interest, givcll sriit;il)lc ~irc,llrl~st;~l~~(ls, iii l)artic~il>ation in talk exchanges that will be profital)le only on the :lssrtmption that they are conducted in general accordance with the CI’ ;III~ the maxims. Whether arty slich conclusion Cil11 l)e reached, I am IIncc:rtain; in any case, I am fairly siirv that I cannot ranch it until I at11 a good deal clearer about the nature of relcvailcc and of tlic c~ircti~ljstances in which it is requirecl. I t i s n o w t i m e t o s h o w t h e conncctioil ljetween tlbe CI’ ;tnd maxims, on the one hand, and conversational implicatllre on the other. A participant in a talk exchange may fail to f’r~lfill a maxim ill various ways, which include the following: 1. He may qrlietly and 1liiostclltatiollsly VIOI,AY*U: a maxim; if so, in some cases he will be lialjle to mislead. 2. He may 0l”‘I’ 0Il’I’ from tlie operatioli botll of thca maxim and of the CP; he may say, indicate, or allow it to l~~~~nie plain that hc is unwilling to cooperate in the- way the maxim reqllires. IIe may say, for example, I c*cllitlot ,sclFg tj~or’e; 1jL{j 1ijl.s (ire scrolled. 3. He may be faced by a (:IASII: He may be unable, for example, to fulfill the first maxim of Quantity (Be as informative as is reqllired) without violating the second maxim of Qtlality (Have adequate evidence for what yo11 say). 4. He may FLOIJT a maxim; that is, he may I3LA’I‘AN’I‘I,Y fail to f’nlfill it. On the assumption that the speaker is allIe to fulfill the maxim and to do so without violating another maxim (hecnttse of a cl;~h), is not opting out, and is not, in view of the blatancy of his performance, trying to mislead, the hearer is faced with a millor prol)letn: I low ca11 his saying what he did say IF reconciled with the srlpposition that he is observing the overall CP? This sitrlation is one that characteristically gives riye to a conversational implicature; and when a conversational implicature is generated in this way, I shall say that a maxim is being EXIXOI’I‘EI). I am now in a position to characterize the notion of colivers~~tioriH1 implicature. A man who, by (in, when) saying (or making as if to say) t h a t 11 has implicated that q, may be said to have convcrsatic~~~~~,lly implicated that (I, PRO\‘II)I~:I) TIIAY’ (I) he is to be presumed to lx; ol)-
so
II. P . Grice
\er\-irlg tllcl collvc~t.s;~tion~~l maxinls, or al le;lst tllc coopcrativc~ principie; (2) the sllpposition tllat hc is aware that, or thinks that, (I is reqriirctl in order to make his s;\yil,g or making as if to say 11 (or doing so in wJ‘JJOSJ< t t r r n s ) CWnsistent with this presumption; and (3) the speaker thinks (and wo111d expect the hearer to think tliat tlie speaker tliinks) that it is within the competence of the hearer to work out, or grasp intllitively, that the supposition mentioned in (2) IS reqrrired. Apply this to my irlitial ~xarnpl~~, to 13’s remark that C has not yet l~rn to prison. In a siiital)lc setting A might reason as follows: ‘(I) I3 lias a p p a r e n t l y vriolatcxcl tire inaxinr ‘I3e relevant’ and so may IW rcbgartlctl as having flo\itcd orre of tire maxims conjoiiring pcrspiclrity, yet I have no reason to srippose that he is optiiig out froin the operation of tlic CP; (2) hrivei tlie circrrinstances, I can regard Iris irrclevrancc as only apparelit if, and ouly if, I sr~ppose him to think that C is potentially dishonest; (3) R k news that I am capable of working out step (2). So H implicates that C is potentially dishonest.’ The presence of a conversational implicatrrre must lx capable of being vvforked orrt; for even if it can in fact he intuitively grasped, ~inlcss the intuition is replaceal)le by an argument, the implicature (if present at all) will riot conlit as a <:ONVF:J~SAI‘JONAI, implicature; it will he a CON\~EIVTJO~NAL implicature. To work out that a particular conversational implicatiire is present, the hearer will reply on the follo\\~iiig data: (1) the conventional meaning of the words used, together n,ith the identity of any references that may be irivolved; (2) the CP anti its maxims; (3) the context, linguistic or otherwise, of the ritterance; (4) other iterns of 1,ackgrorrnd knowledge; and (S) the fact (or slipposed fact) tliat all relevant items falling under the previous Ileadings are available t o both participants and ljoth participants know or asslime this to be the case. A general pattern for the working oiit of a conversational implicature might he given as follows: ‘He has said that 11; there is no reason to suppose that he is not observing the maxims, or at least the CP; he could not be doing this unless he thoright that q; he knows (and knows that I know that he knows) that I c‘au see that the siipposition that he thinks that (1 IS required; he has done nothing to stop me thinking that q; he intends me to think, or is at least willing to allow me to think, that q; and so he has implicated that f~.’ Examples I shall now offer a nrimher three groups.
of examples, which I shall divide into
L o g i c and C o n v e r s a t i o n
A is standing hy an obviously immol~ilized 1,~ B; the following exchange takes place:
51
car and is approached
of petrol. B: There is cl garnge r o u n d t h e corner.
(1) A: 1 urn out
(Gloss: B would be infringing the maxim ‘Be relevant’ unless he thinks, or thinks it possihle, that the garage is open, and has petrol to sell; so he impjicates that the garage is, or at least may be open, etc.) In this example, unlike the case of the remark Ilo llusn’t Brett t o piam yet, the unstated connection ljetween B’s remark and A’s remark is so obvious that, even if one interprets the supermaxim of Manner, ‘Be perspicuolls,’ as applying Ilot only to the expressiotl of what is said l)ut also to the connection of what is said with acljacent remarks, there seems to be no case for regarding that supermaxim as infringed in this example. The next example is perhaps a little less clear in this respect:
(2) A: B:
Smith doesn’t seem to hnve n girlfriend these days. He has been paying (I lot of visits to New York lately.
B implicates that Smith has, or may have, a girlfriend in New York. (A gloss is unnecessary in view of that given for the previous example.) In both examples, the speaker implicates that which he must he assumed to believe in order to preserve the assumption that he is ohserving the maxim of relation.
A is planning with B an itinerary for a holiday in France. Both know that A wants to see his friend C, if to do so would not involve too great a prolongation of his journey: (3) A: Where does C live? B: Somewhere in the South of France. (Gloss: There is no reason to suppose that B is opting out; his answer is, as he well knows, less informative than is required to meet A’s needs. This infringement of the first maxim of Quantity can be explained only by the supposition that B is aware that to he more informative would be to say something that infringed the maxim of Quality, ‘Don’t say
I.ogic
and Conversation
53
11c is (‘Methillks tile 1x1~ ~10th protest too III~IC~I’). 1311t if it is tllor1ght of as designed, it wollld I)e 211 ol)licluc way of conveyiilg tllat it is to SollIe degree cor~troversial whc~tlltlr or not 1,. It is, ljowever, ai.g!rid)lc that such an irnplicxltiire ~011ld 1~ cspl:tir~d by r(~I.(~rcIlcc~ t o tllc m a x i m o f Helation withorlt invoking ~111 a l l e g e d seconc~n on closc~ terms Ilrltil IIOW, has betrayed a secret of A’s to a l)rlsiness rival. A and his atldience l~tll know this. A says ‘X is (I j?luj jhcttd’. (Gloss: It is perfectly ol)viorls to A and his audience that what A has said or has ~nacltb as if to say is s o m e t h i n g h e d o e s n o t ljelieve, ant1 the arldience k n o w s tllat A knows that this is ol,violls to the arldience. So, r~nless A’s rltternnccb is entirely pointless, A must l)cb trying to get across some otllc*r proposition than the one he pllrl,orts t o 1~~ putting forward. ‘I’llis mr~st lx* some 0l)vioiisly related proposition; tlic most ohvioiisly related proposition is the contradictory of the 0116’ l i e p u r p o r t s t o l)e p u t t i n g forward.) 2 . Metcrjhoc. Examples like Yore ure tlw c:t’ecllTr in my cc~ffue characteristically involve categorial falsity, so the contradictory of‘ what the speaker has made as it to say will, strictly speaking, 1~: a trltism; so it cannot he TIIAT that sucll a speaker is trying to get across. WIG> most likely srlpposition is that ‘the spcakt-r is attril)rlting to his arldience some feature or features iI1 respect of which tlicb ati(lienccA resembles (more or less fancifully) the mentioned siihstance. It is possil)le to combine metaphor and iroijy hy imposing on tlte hearer two stages of interpretation. I say You (11.e tile c’~‘(II)~ ill III!/ coffee, intending the hearer to reach first the metaphor interpretant ‘You are my pride and joy’ and then tile irony interpretai1t ‘You are my bane. 3. Meiosis. Of a man known to have broken 111, all tllc> flrrljitrlrc>, one says lie w(~~ (1 little iiztoxiccitd. Every nice girl loves a sailor. 4 . FJyj~erhole. (2l)) Example? in which the second maxim of Quality, ‘Do not say that for which you lack adeqllate evidence’, is flouted are perhaps not easy to find, but the following seems to be a specimen. I Sily of X’s wife, Size is j~ol~ll~lF/ clec~iving hint this cwning. In a sriital)le context, or with a suitable gesture or tone of voice, it nray IW cleat that I have no adeqtlate reason for supposing this to he the case. My partner, to preserve the assumption that the conversational game is still being played, asslImes that I am getting at some related proposition for the acceptance of which I 110 have a rcasonahle basis. ‘The
relatml proposition might wc~ll lx that she is given to dcu:iving her hrlshnd. or possil)ly that slle is tlicb sort of’ ptarson who woltld not stop short of s11ch conclllct. (3)
l<‘sclnlj~lcs ill 1!?1lic*l1
(111 i7)lj,liccltrrre
is clc~llicvctl II!/ rc>cll,
(is tiis-
tillc*f .fro?tl cij’j)trt’c’)rI. I.iolufir~?l 0.f the f)l(lxirn 0.f Ilclotiorr are perhaps rare, l)llt the t’ollowillg scums to IX> a good candicI;ltc~. At ;\ genteel tea party, A says A1r.s. S is (1r1 oltl Iwig. ‘I’hc~re is a inoint~lIt of appalled silflllc’(l, ;Illd then 1 3 s;i)‘s Tllc~ roctrillcr I1tr.r Iweri qftitcJ tl(,/i,glitsfitl Illis .PI/~~I~)~o~, Ill/.cr~‘! it.2 13 lli\s 1,latantly reftlsccl to make wh;lt 1114; says relej\.ant to A’s prc(scclir\g remark. I Ic therel,y implicates tllnt A’s remark sho~lld Ilot l)e tliscilssed alld, p e r h a p s more specifically, that A has committed a social gaffe. ( 4 ) Estririjr1e.s it1 tc;liic*l1 wrriorts maxims .ftrlling rrndrzr tile sujwrn1uxinl ‘Re j,ersj,ic:rcorc.r’
trre j7orited
We must remember that we are concerned only with ambiguity that is deliljerate, and that the speaker intends or expec*ts to l)e recognized 1,~ llis hearer. The problem the hearer has to sol\~e is why a speaker shol~ld, when still playing the conversational game, go orlt of his way to choose an amhigllous rltterance. There are two types of cases: (a) Examples in which there is no difference, or no striking difference, 1Mween two interpretations of an utterance with respect to Stl-ilifilltfO1.Wi~r(ll~cSS; neither intcrpretatiolb is notal,ly more sophisticated, less stalldard, more recondite or more far-fetchc~d than the other. We might consider Blake’s lines: ‘Never seek to tell thy love, Love that never told can be. To avoid the complications introduced 1)~ tile presence of the imperative mood, I shall consider the related sentence, I .sought to tell TTZI/ love, love tht newr told c*(I~~ be. There may he a double ambiguity here. My love may refer to either a state of emotion or an object of emotion, and love that never told curl he may mean either ‘Love that cannot be told’ or ‘love that if told cannot continue to exist.’ Partly because of the sophistication of the poet and partly lIecause of internal evidence (that the ambiguity is kept IIP), there seems to 1~ no alternative to supposing that the ambiguities are deliherate and that the poet is conveying both what he would be saying if one interpretation were intended rather than the other, and \,ice versa; though no doubt the poet is not explicitly SAYING any one of these things hut only conveying or suggesting them (cf. ‘Since she [nature] pricked thee out of women’s pleasure, mine he thy love, and thy love’s use their treasure.) (1)) Examples in which one interpretation is notably less straightforward than another. Take the complex example of the British Gen1.
r\nll)ig!rrity.
Logic arid
Conversation
55
era1 who captured the town of Sind an<1 sent hack tlje mcssagc Z’ec*cscl~i. The aml,igqli ty involved (‘I have Sind’/‘I havc~ sinnt4’) iq phonemic, iiot morphemi<~; and tlie expression actually tiscd is unamhigliolis, l)ut since it is in a langllage foreign to speaker and hearer, translation is called for, and the ambiguity resides in the stailtlard translation into native English. Whether or not the straightforward interpretant (‘I have sinned’) is being conveyed, it seems that the rlonstr;li~glltforwar
Miss X .\ung ‘Iiome sweet home. Mi.v.v ,Y ~~~*odr~cc~l (I .vc~rio.v 0.f .vorl,rtl.v tllcl t c.o~r.cl.~l’orlcl~~(~ chewy with the score of’ ‘Home sfveet home’.
1 1\;1\,(> s o l;\r c~oiisitl(~rc~tl otjly c’;iscs o f wllat I iijigllt (311 I);irticrllarizc~cl collvcrs;~tic,ilal iiilplicatllre- that is to say, cast’s in which an irnl)lic3trir~~ is carric~cl l,y saying Illat It on a pnrticrllar occasion in virtrI(a 01‘ s;l)c~(~ial f;batrlrc>s of the context, cases in which thcrt: is no room fi)r tllc id(ta tllat ali inil)lic3tllrc~ of this sort is NOl\hlAl.l,Y carried l)y saying tliat 1). R\lt there are cases of g!cneralized conversational iniplicatllrc. Somc~tiinc~s orit’ can sa) that tlic 11s~’ of a certain form of WortIs in a11 Ilttt3r;llice worlld IlormaIIy (in the AI3SEN(:F; of Special circllmstanc~cs) carry sl~c~li-~~nd-s~~<~li a11 implicature or type of implicatllrca. Nonc,olltrovt~rsinI examples are perhaps hard to find, since it is all too easy to trtbat a gt3lit~ralized conversational implicntllre as if it urerc a coli\~cntioiial implicature. I offer an example that 1 hope may 1~3 fairly Iionco1i troversial. An)~onc \~~I10 lists a sentcnc~t~ of t h e f o r m X i s ntceli71g f/ iuo7no7i thin cr’f’ttirlg \\‘0111d riorn~ally implicate that the person to he met was som~~onc other tllarl X’s wifcb, motlier, sister, or perhaps even close platoilic fricrid. Similarly, if I wore to say X luztlt illlo (1 /louse yesterrlug utrtl ,fi)r/t~tl (I tortoise irrsitle ~ltc~frc~nt floor, my hcart’r would normally l)e stlrprisul if some time later I revealed that the ho~lse was X’s ow11. I collld proclrlce similar linguistic phenomena involving the expressions (1 g(It’~/ctl, 0 (*f/r, (1 c*ollcge, and so on. Sometimes, howe\rer, tlierc woi~ld normalIy 1~2 no such implicatrlre (‘I have been sittillg in a car all morning’), and sometimes a reverse implicature (‘I l,rokc a finger >vesterday’). I am inclined to think that one would not lend a sympatlletic ear to a philosopher who suggested that there are three qenses of the form of expression (I)? X: one in which it means ro11ghIy ‘somethirlg that satisfies the conditions defining the word X,’ anotlic~r iii which it mealis approximately ‘an X (in the first sense) that is only remotel!~ related in a certain way to some person indicatetl by the context,’ a11cI )ret another in which it means ‘an X (in the first sense) that is closc~ly related in a certain way to some person indicatc~(l 1)y the csontc>xt.’ W011ld we not m~icli prefer an account on tlic following lines (whicll, of coilrse, may l)e incorrect in detail):
Logic
and
Conversation
Fi7
When someone, by rrsing tlrc form of’cxprr~ssiorr (111 X, implicates that the X does not belong to or is not otherwise closely c+onnectd with some ideiltifial,le person, tire implicatrirc i s p r e s e n t l~cause the
I
speaker has failed to IX specific in a w a y irr wlrioh Ire might hnvc~ been expected to 1~ specific, with the consequence that it is likely to he assumed that he is not in a position to 1~ specifirm. This is a familiar implicnt1ire situation and is classifialrle as a fkilrirc, for one reason or another, to fulfill the first maxim of Qiantity. The only tliffiorilt question is why it sliould, in certain cases, be presumed, independently of information alrorrt pnrtictrlar c o n t e x t s o f rrtteranee, t h a t specification of the closeness or remoteness of’ the connection between a particular person or ol)ject and a ftrrther person who is mentioned or indicated 1)~ the utterance slrould 1~ likely to Ire of‘ interest. The answer must lie in tire following region: Transnctiotrs between a person and other persons or things closely connectc!d with him are lialjle to l)e very different as regards their concomitants and results from the same sort of transactions involviirg only remotely connected persons or things; the concomitants and results, for instance, of my finding a hole in MY roof are likely to be very different frorn the concomitants and results of my finding a hole in someone else’s roof. Information, like money, i s o f t e n g i v e n withorit the giver’s knowing to joist what rise the recipient will want to put it. If someone to whom a transaction is mentioned gives it further consicleration, he is likely to find himself wanting the answers to ftrrther questions that the speaker may not he able to identify in advance; if the appropriate specification will lx likely to enable the hearer to answer a considerable variety of such questioris for himself, then there is a presumption that the speaker should incliide it in his remark; if not, then there is no such presumption. Finally, we can now show that, conversational implicature Iieing what it is, it must possess certain features: 1. Since, to assume the presence of a conversational implicature, we have to assume that at least the Cooperative Principle is being observed, and since it is possible to opt out trf the observation of this principle, it follows that a generalized conversational implicatiire can he canceled in a particular case. It may be explicitly canceled, lry the addition of a clause that states or implies that the speaker has opted out, or it may be contextually canceled, if the form of utterance that usually carries it is used in a context that makes it clear that the speaker IS opting out. 2. Insofar as the calculation that a particular conversational implicature is present requires, hesides contextual and background infor-
54
JJ. P. Grice
O f ’ \Vllklt IILLS I)C?ell s a i d (01. O f tIlf2 C!O~lVf?Jltioilal conllnitmcirt of the Iitter;ince), and insofar as the manner of expression plays i10 role in the calclilation, it will not he possible to IlliltiOll, Ollly il kllOWIC‘d~~L
. f i n d anothc~r w a y o f sayirrg tllo same’ t h i n g , wl~ich siml)ly lacks the irnplicatrire ii) c4\iestion, extbept wllcrc SOJTI~ special feattire o f t h e ulil,stitlitecl version is itself rele\pant to the determination of an implicatI1r(b (in virtlle of ant’ of the> rl,axinjs of Manner). II‘ wt’ CXII this featiirc NoSl~l~:~l‘~~~:lil\l~ll .I’I‘l‘, OIIC m a y e x p e c t a generalizc~tl cbonversational iinpli(*afiir(~ that is cnrrirY1 I)y a fainiliar, nonspecial locrltion to II;l\‘V ;1 IliglI (l('~l~(~(~ Of Il~~l~(l(~ti~(‘llitl)ility. 3. ‘I‘0 Sj)(‘i\k HPProsii~~~~tt~ly, since tl\cb calc*lll;\tion of‘ tl\c prc’scbnccb of a ~ol~~‘(~rsi~tionnl implic~wtl~rc~ prcsripposes an initial k~~owl(~clg!c of the c~onvc~ntional force of the expression the utterance of which [harries the implicatllre, a conversational implicatum will 1~ a condition that is ijot iilcllicled ii) the original specification of tile expression’s c!onventional force. Tho~~gh it may not IX impossible for what starts life, s o t o speak, a s a conversational implicature to l)ecome conventionalized, to siippose that this is so in a given case wolild require sptc*ial jIIstification. So, initially at least, conversational implicata are ijot I>iirt of the meaning of the expressions to the employment of wliich they attach. 4. Since the tnitli of a conversational implicaturn is not recluired 1)~ the trlltll o f w h a t i s s a i d ( w h a t i s s a i d m a y 1)e trlie-what is iinplicatetl il~;\y IW false), the iniplic*nture is not carried 1)y what is said, l,rit onl~~ l)y the saying of wllat is said, or l)y ‘plltting it that way.’ Fj. Since, to calcrilate a coi~\lersr~tionul implicatlire is to calculate what leas to 1~ slipposed in order to preserve the s\ippositionthat the Coopc~rntivc P r i n c i p l e i s l)eing ol)served, and since tlltre m a y he \rariolis possible specific explanatioils, a list of’ which mny 1x2 open, the conversational implicatuni in such cases will IW disjlinction of such specific explanations; and if the list of these is open, the implicatum w i l l 1lave j u s t t h e k i n d o f i n d e t e r m i n a c y t h a t m a n y a c t u a l implicata do in fact seem to possess.
